When blood glucose concentrations rise (e.g., in response to meals), amylin is co-secreted with insulin from the cells while glucagon secretion from the a cells is inhibited. Conversely, as glucose levels decline, insulin and amylin secretion falls while glucagon secretion increases. Physical exercise is also accompanied by falling glucose, falling insulin and rising glucagon concentrations. In contrast, in Type 1 diabetes, due to concomitant a cell dysfunction, there is hyperglucagonemia in the postprandial state and inadequate glucagon response during exercise thereby predisposing to postprandial hyperglycemia and post exercise hypoglycemia respectively. Recent studies with closed loop artificial pancreas systems have applied single-hormone (insulin only) or dual hormone (insulin + glucagon) approaches for short term glucose control in type 1 diabetes without attempting to restore normal hormonal milieu in these individuals during meals and exercise. The innovative approach outlined in this application seeks to restore, to the extent possible, normal hormonal milieu with meals and physical exercise. Using dual-chamber pumps and imminently available co-formulation of insulin and the amylin analogue pramlintide as a control mechanism, we aim to recreate normal postprandial hormonal profile in type 1 diabetes participants. The use of pramlintide will help suppress postprandial hyperglucagonemia, improve insulin sensitivity thus permitting a lower prandial dose of insulin. That in turn should prevent late postprandial hypoglycemia. The use of glucagon will be limited only as a safety-rescue mechanism that will be triggered by declining glucose levels especially during and after physical exercise. We also propose to apply the computed insulin sensitivity parameters (with pramlintide and during exercise) using the insulin pump infusion rates and CGM derived glucose excursions to inform the algorithm, test its safety in silico experiments, refine the tri-hormonal (insulin + pramlintid + glucagon) algorithm and eventually test it's efficacy in a long-term clinical trial vs. single- (inulin only) and bi- hormonal (insulin + pramlintide) algorithms. The proposed studies therefore incorporate several innovations that include the following: application of insulin sensitivity parameters during closed loop studies; use of insulin-pramlintide as a control mechanism to mimic the healthy cell; amelioration of postprandial hyperglucagonemia with use of insulin-pramlintide combination; use of glucagon only as a rescue mechanism for impending hypoglycemia especially during exercise as occurs in health thus mimicking the healthy a cell; finally and most importantly testing the tri- hormonal approach versus contemporary single- and bi-hormonal systems in a long term at home clinical trial.